2450 ml of ether and 500 ml of water. The ether was removed under reduced pressure to yield a solid which on crystallization from methanol yielded 14 g (84Z) of product, mp 75-76"; [a]"D -10" (c 1.0, methanol). The (3=) nitrile gave mp 65-66"; nmr (CC14) 6 7.25-7.1 (m,10 H), 3.92 (d, 1 H, J = 9.3, 3.33 (m, 1 H), 1.21 (d, 3 H, J = 7.0). Anal. Calcd for C18H16N: C, 86.84; H, 6.83; N, 6.33. Found: C,86.60; H, 6.91; N,6.59. The l-deuterio-2-methyl-3,3-diphenylpropionitrile was prepared by exchange in methanol-d with 1.0 N sodium methoxide, nmr (CC14)67.2~7.1(m,10H),3.91(s,1H),1.19(s,3H).
l-Tri~o-l-methyl-2,2-diphenylcyclopropylnitrile. To an ether solution of 2.0 g (0.009 mol) of 2,2-diphenylcyclopropylnitrilewas added an equivalent amount of lithium diisopropyl amide dissolved in ether. The deep red solution was hydrolyzed with 1.0 ml of tritiated water (250 WCijml) to yield 1.2 g of nitrile with 114 dec/ ~nin/mg.~OTable VI1 is typical of the rate data for tritium-hydrogen exchange. Straight lines were obtained in all cases and the average of ten runs gave an average slope of 8.0 X min-l. (30) Scintillation counting was performed using PPO (4 g/L) and POPOP (50 mg/l.) in toluene-counting solution in conjunction with a Packard Tricarb liquid scintillation counter.
Manganese (111) Complexes in Oxidative Decarboxylation of Acids James M. Anderson and Jay K. Kochi
Contribution from the Departments of Chemistry, Case Western Reserve University, Cleveland, Ohio, and Indiana University, Bloomington, Indiana 47401. Received August 11, 1969 Abstract: Manganese(II1) effectsoxidative decarboxylation of a variety of acids in nonaqueous solutions. The products and stoichiometries of the decarboxylation are examined. The reduction of MnIII follows first-order kinetics. Autoretardation by the MnII formed in the reaction is attributed to mixed valence complexes between MnIII and Mn". Alkyl radicals (and carbon dioxide) formed by multibond homolysis of the MnIII carboxylates are oxidized by a second Mn"1 to alkenes and esters. The enhanced rate of decarboxylation and oxidation of alkyl radicals by MnIII in the presence of strong acids is ascribed to cationic Mn"1 species. Copper(I1) effectively traps alkyl radicals from the decarboxylation. The autoxidative decarboxylation of pivalic acid in the presence of oxygen is catalyzed by Mn"1 and produces high yields of t-butyl alcohol and di-r-butyl peroxide.
variety of transition metal compounds have been employed to catalyze the autoxidation of hydrocarbons. In many cases, these autoxidations involve a rather complex set of reactions and metastable intermediates. In particular, the role of free radicals and their interaction with the metal species are not clear. Manganese(II1) complexes have been used t o oxidize a number of types of organic functional groups.2 When manganese compounds are employed in a catalytic capacity, the metal species is thought to alternate between the I1 and I11 oxidation states. MnlI1 in aqueous solutions, however, is especially prone to disproportionate (eq 1). The latter does not
mediacy of alkyl radicals. The abilities of MnlI1 and Cu'I to oxidize alkyl radicals are also compared.
A
2MnI11
e MnII + MnIV
(1)
appear t o be as severe a limitation in nonaqueous solutions, since MnlI1 complexes are relatively stable in these media. In this report, we demonstrate the use of M P complexes in the oxidative decarboxylation of acids. Products and kinetic studies are coupled in order to clarify the mechanism of the oxidation, the catalysis by strong acids, the retardation by Mn'I and the inter(1) For a review, see (a) N. M. Emanuel, E. T. Denisov, and Z. K. Maizus, "Liquid Phase Oxidation of Hydrocarbons," Plenum Press, New York, N. Y.,1967; (b) "Oxidation of Organic Compounds," Advances in Chemistry Series, No. 76, Vol. 2, American Chemical Society, Washington, D. C., 1968; (c) Angew. Chem. Intern. Ed. Engl., 8.97 (1969). (2) W. A. Waters and J. S.Littler, "Oxidation in Organic Chemistry," K. B. Wiberg, Ed., Academic Press, New York, N. Y., 1965, Chapter 3. (3) Except where it is pertinent to the discussion, the ligands associated with the manganese species will not be included. Octahedral coordination generally pertains.
Journal of the American Chemical Society
Results All oxidative decarboxylations by Mn"' were studied thermally in homogeneous and degassed solutions. Manganese(II1) carboxylates were generated in situ (eq 1) by dissolving manganese(II1) acetate in the carboxylic acid as solvent. The products, the stoichiometry, and the kinetics of the oxidation of three acids, chosen t o be representative of tertiary (pivalic), sec-
+
M~I~~~(OAC 3RC02H )~
e Mnrrr(02CR)3+
3HOAc (2)
ondary (isobutyric), and primary (n-butyric) acids, were examined in detail. Qualitatively, the rates of reduction could be followed visually, since the color of the solution changed from deep burgundy to colorless as MnlI1 was converted to Mn". Products and Stoichiometry of Oxidative Decarboxylation by Mn"'. Pivalic Acid. The oxidative decarboxylation of pivalic acid was carried out in solutions of the neat acid containing 4% by wt acetic acid t o maintain fluidity and facilitate handling at room temperature. From each mole of Mn"', 0.5 mole each of carbon dioxide and isobutylene were formed together with smaller amounts of t-butyl pivalate and acetate. Excellent material balances based on eq 3 (where 2Mn"I
+ (CH3)3CC0SH+2Mn" + [t-Bu+],. + H+ + COS (3)
[t-Bu+Iox= (CH3)&=CH2, ~ - B U O ~ C C ( C H etc.) ~ ) were ~, obtained (Table I).
1 92:8 April 22, 1970
245 1 Table I. Decarboxylation of Pivalic Acid by Mn(II1) at 125" Products, mmoles
c
Mn(III),b mmoles 2.72 2.76e 2.77 2.76 2.76O 2.80 2 . 74e 2.74 2.71 2.84 2.76 2.77 2.80 2.70 2.78 2.72 2.78
Add./ Reaction Mn(II1)" time, min
Additive
Cu(0Ac)z Mn(OAc), Mn(OAc)* Mn(OAc)z Mn(OAc), TFAf TFAf TFA,f CuIr 9 TFAf TFAf TFA,fCurl9 H2SO4 HzSO4 HC104 HClOd
0.05 1 .o 1 .o 2.0 2.0 1.o 1.0 1.0 2.0 3.0 3.0 1.0 3.0 1 .o 3.0
90 120 90 150 150 150 150 15 10 15 10 10 10 20 15 10 10
2COd COZ 1.37 1.37 1.40 1.38 1.40 1.42 1.40 1.35 1.36 1.34 1.39 1.37 1.32 1.26 1.29 1.33 1.33
Mn(II1)" i-C4HlO 1.01 0.99 1.01 1.00 1.01 1.01 1.03 0.99 1.01 0.95 1.01 0.99 0.95 0.93 0.93 0.98 0.96
0.01 0.02 th
0.04 0.02 0.05 0.02 t t
0 t t
0 t t t t
a In 10 ml of pivalic acid-acetic acid (96:4 wt/wt). * Mn(OAc),.2HZ0. In 10 ml of pivalic acid-acetic acid (80:20 v/v). f Trifluoroacetic acid. (250 25 60 21 58 >250 >250 >250 >250 >250 >250 >200 >150 >120
0.80 0.84 0.93 0.75 0.85 0.75 0.85 0.87 0.85 0.88 0.84 0.83 0.83 0.81 0.64 0.50 0.08
Mrnoles/mmole. d t-Buox includes isobutylene and esters. Trace amounts CU(OAC)~ based on Mn(II1). 5 mole
+ (CH&CHCO,H + 2Mn"
+ [i-Pr+Iox+ COZ+ H+
(7)
= C3Hs, i-Pr02CCH(CH3)2, etc.). The effects of such additives as sulfuric acid, pyridine, and lithium acetate are also given in Table 11. n-Butyric Acid. The reduction of Mn'" in n-butyric acid proceeded slowly at 125" and produced carbon dioxide and propane in poor yields, together with minor amounts of propylene, n-propyl, and isopropyl butyrates (Table 111). The rate of Mn"' reduction and the amount of decarboxylation were markedly increased by the addition of trifluoroacetic acid. Similar effects were shown by sulfuric and perchloric acids. The generally poor material balances for the decarboxylations of n-butyric and isobutyric acids by Mn"' were attributed to side reactions involving nondecarboxylative processes. Hydrogen loss from the a-position of the acid to form carboxyalkyl radicals is one such oxidation which has been carefully delineated with acetic acid.6-8 MnrrI
+ CH3COzH
---t
.CHzCOOH
+ Mnrl + H+
(8)
Competitive Decarboxylation of Acids by MnIX1. Mixtures of pivalic acid and n-butyric or isobutyric acid were decarboxylated with Mn"' in the absence and presence of varying amounts of trifluoroacetic acid (Table IV). Yields of carbon dioxide, alkane (isobutane, propane), and alkene (isobutylene and propylene) were measured quantitatively. The yields of the accompanying esters were obtained by simple extrapolation of the detailed data for the decarboxylation of the separate acids presented in Tables I, 11, and 111. The validity of this extrapolation was cross checked with competition studies in the presence of CuII, since decarboxylations under the latter conditions generated only alkene (videsupra). (6) R. E. van der Ploeg, R. W. de Korte, and E. C. Kooyman, J. Catal., 10,5 2 (1968). (7) E. I. Heiba, R. M. Dessau, and W. J. Koehl, Jr., J. Am. Chem. SOC.,91, 138 (19693; 90, 5905 (1968); Discussions Faraday SOC.,46, 189 (1968). (8) See, however, J. B. Bush, Jr., and H. Finkbeiner, J . Am. Chem. SOC.,90, 5903 (1968); Discussion Faraday SOC.,46,150 (1968).
Anderson, Kochi
/ Manganese(III) Complexes
2452 Table 11. Decarboxylation of Isobutyric Acid by Mn(II1) at 115"
Products, mmoles Mn(II1): mmoles 2.66 2.70 2.65 2.72 2.72 2.70 2.72 2.64 2.72 2.78 2.72 2.73 2.62 2.65 2.65 2.62 2.70 2.71 2.71 2.67 2.70 2.69 2.70 f
Add./ Reaction Mn(I1I)c time, min
Additive
Cu(0Ac)z Mn(OAc)z Mn(OAc)z Mn(OAc)z TFAe TFAe TFAc TFA,e CUI] f TFAe TFA6, CuT1f HzSOa &SO4 HzS04 His04 CsHbNO C~HSNO C SHEN' LiOAc LiOAc LiOAc LiOAc
300 300 300 880 1020 990 120 120
0.5 0.5 1.0 2.0 0.5 1 .o 2.0 1.0 3.0 3.0 2.1 2.0 1.4 1.4 0.5 1.o 2.0 1 .o 1 .o 2.1 2.6
90 120 60 60 30 30 30 30 760 780 780 180 240 180 120
C02
2co2/ Mn(II1)C
C3Hs
C&
i-PrOzCi-Pri-ProxC,d/ i-Pr O Z C C H ~ CSHS
1.28 1.28 1.12 1.29 1.29 1.30 1.34 1.34 1.42 1.14 1.31 1.24 1.48 1.60 1.50 1.56 1.23 0.91 1.34 1.26 1.27 1.01 0.98
0.90 0.95 0.85 0.95 0.95 0.96 0.99 1.02 1.04 0.82 0.96 0.91 1.13 1.21 1.13 1.19 0.93 0.67 0.99 0.94 0.94 0.75 0.73
0.55 0.57 th 0.63 0.62 0.63 0.57 0.47 0.38 t 0.59 0.11 0.08 0.07 0.07 0.10 0.60 0.61 0.67 0.56 0.65 0.64 0.66
0.15 0.15 0.85 0.09 0.05 t 0.23 0.34 0.43 0.88 0.45 0.86 0.32 0.42 0.49 0.55 0.11 0.10 0.07 0.08 0.10 0.07 0.08
0.13 0.12 t 0.11 0.07 0.02 0.19 0.28 0.23 0.03 0.30 0.11 0.57 0.71 0.38 0.42 0.17 0.12 0.10 0.12 0.11 0.09 0.09
0 0 0.04 0.11 0.05 0.06 0.07 0.12 0.06 t t t 0.08 0.11 0.11 0.11 0 0.04 0.07 0.05 0.07 0.09 0.05
a In 10 ml of isobutyric acid. Mn(OAc)s.2H20. Mmoles/mmole. i-Prox includes propylene and esters. 5 mole Cu(0Ac)z based on Mn(II1). 0 Pyridine. * Trace amounts (250 0.5 0.3 0.1 0.8 1.6 1.9 >250 1.3 8 12 18 13 11 0.5 0.4 0.4 0.5 0.4 0.4 0.3
Zi-Prc/ COZ 0.65 0.66 0.80 0.73 0.61 0.55 0.79 0.90 0.77 0.80 1.02 0.87 0.71 0.82 0.70 0.76 0.72 0.94 0.68 0.64 0.73 0.88 0.90
Trifluoroacetic acid
e
Table 111. Decarboxylation of n-Butyric Acid by Mn(II1) at 125" ~~
____
____
_
_
_
_
~
~~
-Products, mmoles Mn(III),b mmoles 2.77 2.78 2.79 2.92 2.73 2.77 2.74 2.72 2.72 2.75 2.77 f
Add./ Reaction Mn(II1)Ctime, min
Additive CU(0Ac)z Mn(OAc), Mn(0Ac)Z TFAe TFA," Cu" TFA* TFAe, Cu" HzSOI HzSOa HClO4
f
f
0.05 1.o 2.0 1 .o 1.0 3.0 3.0 1 .o 3.0 1.0
720 900 780 780 180 240
60 90 60
60 60
COZ
2COd Mn(II1)o
C3Hs
C3Hs
0.67 0.50 0.58 0.64 0.79 0.73 0.93 0.94 0.83 0.78 0.76
0.48 0.36 0.42 0.43 0.58 0.53 0.68 0.69 0.61 0.57 0.55
0.27 t 0.19 0.20 0.43 0.02 0.50 0.02 0.25 0.10 0.27
0.04 0.19 0.03 0.03 0.06 0.50 0.09 0.67 0.10 t 0.11
In 10 ml of n-butyric acid. Mn(OAc)a.2H20. Mmoles/mmole. 5 mole CU(OAC)~ based on.Mn(II1). 0 Trace amounts (70 0.6 >28 1.4 >80 1 .o >40 3.0 >150
13 128 18 >150 31 >140 13 >170 12 >200 75 >130
17 28 28 21 16 25 25 21 23 25 24 11
1.28 1.46 1 .00 0.97 1.42 1.40 1.11 1.06 1.55 1.48 1.33 1.33
0.005 0 0.03 t 0.01 0 0.05 t 0.009 0 0.07 t
B = n-Butyric Acidi 0.009 0.05 0.02 t 0.05 0.02 0.04 t t 0.01 0.02 t 0.01 t 0.04 t t 0.007 0.014 0 0.02 t 0.09 t
0.48 0.82 0.31 0.43 0.83 0.90 0.51 0.51 0.83 0.76 0.65 0.68
4 >4 1.3 >8 4 8 4 18
13 >190 10 >lo0 >100 >220 >120 >120 >150 >200 >170 >180
240 460 400
650
500 530 490 900 1040 500 500 500
In solutions containing m. 2.7 mmoles of M ~ ( O A C )2Hz0 ~ . in 10 ml of carboxylic acid (component A = pivalic acid and component B = isobutyric or n-butyric acid). b Moles/mmole. c Trifluoroacetic acid. C3Hl+ includes propylene and propyl esters; C4Hg+includes isobutylene and [-butyl esters. e To closest significant figure. f ZCa includes all propyl products; ZCCincludes all butyl products. 0 105". 5 mole CU(OAC)~ based on Mn(OAc)r.2H~0added. 120". j Trace amounts. (I
retarding effect exerted by MnT1in relation to its concentration. For example, when a 5 molar excess (relative to Mn") of trifluoroacetic acid was employed, the disappearance of Mn"' followed first-order kinetics to beyond 90 reduction.
The Effect of Metal Carboxylates. The effect of various metal carboxylates on the rate of reduction of Mn"I by pivalic acid is given in Table VIII. These 1.0
0.8 1.0
I
0.6
F 0.4
..
I
pd
0.4
9
\
5
I
9
\
8
0.6
9
\ \
0
I 0
P, m
\\
0
-1
0
,
,
0
.\\\;\.
\\
0.21
'.
"
3 0.2
0
0
0
\
0.1 0
10
20
30
40
50
Time, min.
0.1
0
20
40
60
80
100
Time, min.
Figure 1. Kinetics of Mn(II1) reduction in pivalic acid at 80". Effect of MnIII: 0 , 2.60 X M M Mn(0Ac)s; 0, 2.60 X M ~ ( O A C )5.26 ~, X MMn(0Ac:Z.
Solvent Isotope Effect. The deuterium isotope effect was studied by comparing the reduction of Mn"' in pivalic acid and pivalic acid-d under equivalent conditions. The reduction of Mn"' was 1.3 f 0.1 times faster in the protic medium at 70".
Figure 2. Kinetics of Mn(II1) reduction in pilvalic acid at 70". Effect of trifluoroacetic acid: 8 , 2.60 X M Mn(0Ac)s; 0, 2.60 X loea M Mn(OAc)s, 2.63 X 10-3 M Mn(OAc)t, 13.82 X 10-3 M TFA; q 2 . 6 0 x M Mn(OAc),, 13.82 x M TFA.
first-order rate constants (k,) are compared with that kl obtained in the absence of additives in the last column of the table. The effects of metal carboxylates on the reduction of Mn"' by pivalic acid were also studied in the presence of added trifluoroacetic acid. Under these conditions, the metal carboxylates underwent metathesis according Anderson, Kochi / Manganese(ZII) Complexes
2454 Table VII. Effect of Trifluoroacetic Acid on the Reduction of Mn(II1) in Pivalic Acid" C F ~ C O ~ H103 , M
CFICOIH/ Mn(I1I)b
Od
06
2.76d 2.76' 5 . 52d 5.52
8.3d 8.36 11.04d 11.04c 13.82d 13.82e
z
1.05 1.05 2.11 2.11 3.17 3.17 4.23 4.23 5.28 5.28
4.3 f 0.2 5.0 =t0 . 5 7.5 i 0 . 1 7.8 f 0 . 4 13.1 f 0 . 1 11.1 f 0 . 7 14.9 f 0 . 3 14f 1 19.5 f 0 . 6 20 f 1 28.9 & 0 . 2 27 f 1 ~~
In 99 wt pivalic acid-acetic acid solutions containing 2.60 X 10-3 M manganese(II1) trisacetate at 70". Molar ratio. Average of at least two runs. d Dihydrate. Anhydrous. a
Wavelength, nm
Figure 3. Visible absorption spectra of Mn(OAc), 2H20 with added Mn(OAc)s in 80 vol 2 pivalic acid-acetic acid: -,2.58 X MMn(III), 5.20 X M MMn(II1); - - - - - -,2.58 x M MnMn(I1); . . . . . . . . ., 2.58 X 10-3 M Mn(III), 7.80 X M Mn(III), 13.0 X low3M Mn(I1); , 2.58 X (11); -.-.- - _ ,2.58 X M Mn(III), 20.8 X MMn(I1).
-
to eq 10. Since trifluoroacetic acid is significantly M"+(OAc),
+ nCF3C02H e MnC(OzCCF3), + ~ H O A C(10)
stronger than acetic acid, we assumed that the free trifluoroacetic acid (CF3C02H)effremaining after neutralization was given by eq 11, where (CF3C02H),was (CF3C02H)eff
=
(CF3COzH)o - n[M(OAc)n] (1 1)
Table V. First-Order Rate Constants for the Reduction of Mn(III)a kl: 104 sec-1
Mn(III), l o 3 M 5.22~ 5 . 20d 3.92~ 3.90d 2.61~ 2.60d 1.3lC 1 .30d
8.6 f0.3 10.0 i 0 . 1 8.4 f 0.3 9.2 f0.2 8.9 f 0.6 9.7 f0.5 8.3 f0.2 9.9 i 0.4
5z
z
In 99 wt pivalic acid-acetic acid solution at 80". of at least two runs. c Dihydrate. d Anhydrous.
Average
Table VI. Effect of Mn(L1) on the Reduction of Mn(II1) by Pivalic Acida Mn(II1): 103 M
Mn(II),e lo3 M
2.611 2.600 2.62, 2.60g 2.62, 2.60 2.62, 2.60s
0 0 1.32 1.32 2.64 2.63 5.28 5.26
Mn(II)/ Mn(II1)d
kl,e 104 sec-1
8.9 f0.6 9.7 f0.5 4.24 f 0 . 0 3 3.3 0.2 2.94 f 0.02 2.37 f 0.07 1 . 8 f 0.1 1 . 9 i 0.1
0.5 0.5
*
1.o
1.0 2.0 2.0
the amount of acid charged and n was the oxidation state of the metal. The experimental first-order rate constant was corrected for the amount of trifluoroacetic acid lost by neutralization according to (eq 9 and 11). The corrected rate constant k, was compared with the experimentally determined rate constant kM in the presence of various metal carboxylates. The validity of this assumption was demonstrated by comparing the effects of lithium and sodium acetate with lithium and sodium trifluoroacetate at corrected concentrations of trifluoroacetic acid. A slight positive kinetic salt effect was noted in these cases. The results in Table VI11 show that Mn" and Co" are unique in their retarding effect on the reduction of Mn"' by pivalic acid in the presence as well as in the absence of trifluoroacetic acid. Activation Parameters. The temperature dependence of the first-order rate constant for the reduction of anhydrous Mn"' trisacetate and its dihydrate in pivalic acid was the same within experimental error. These results are found in Table IX, together with those obtained in the presence of trifluoroacetic acid. The Absorption Spectra of MnT" and MnT1Systems. The visible absorption spectrum of manganese(I11) trisacetate in pivalic-acetic acid solutions showed an absorption maximum at 462 nm ( E 360 M-' cm-I), which followed the Beer-Lambert relationship to within in the concentration range 3 X to 3 X M . Manganese(I1) bisacetate showed no appreciable absorption in this region. The absorption maximum of 2.58 X M Mn"I at 462 nm diminished with increasing amounts of added Mn'I (Figure 3), and at high ratios of Mn"/ MnlII, a new maximum appeared at 494 nm ( E 250 M-' cm-l, MnT1 = 1.30 X M ) . No clear isosbestic points were noted. However, when a solution containing 2.58 X M Mn"' and 2.08 X M Mn" (Le., Mn"/Mn"' = 8) was diluted, the apparent extinction coefficient in the region between 450 and 500 nm increased. Finally, at the dilution of MnlI1 of less than M , Mn" (lou3 M ) exerted no noticeable effect on the absorption spectrum of Mn"'. The effect of MnlI on the spectrum of Mn"' at a given concentration also increased with decreasing teniperatures. A variety of other metal carboxylates including lithium, sodium, and magnesium acetates and trifluoro-
ManIn 99 wt pivalic acid-acetic acid solution at 80". ganese(II1) trisacetate. c Manganese(I1) bisacetate. d Molar ratio. * Average of at least two runs. Dihydrate. Anhydrous. (I
@
Journal of the American Chemical Society 1 92:8 J April 22, 1970
2455 Table VIII. Kinetics of the Reduction of Mn(II1) at 70".a-c The Effect of Salts as Bases Added salt
Added salt, M x 103
LiOAch LiOAci NaOAch NaOAc' Mg(0Ac)ah Mg(OAc)ah Mg(0Acy Mg(OAc)2' Cu(OAc)ah Cu( OAcW' Mn( 0Ac)lh Mn(OAc)zh Mn(OAc)$ Mn(0AcP Co(OAc)ah Co(OAc)ah Co(0Ac)z' Co(0Ac)l' LiTFAh LiTFAi NaTFAh NaTFAi
2.54 2.54 2.58 2.58 2.61 2.61 2.61 2.61 2.62 2.62 2.63 2.63 2.63 2.63 2.63 2.63 2.63 2.63 2.5 2.5 2.5. 2.5
TFA,$ x 103
M
13.2 13.2 13.2 13.2 13.8
0 13.8
0 13.2 13.2 13.8
0 13.8 0 13.8 0 13.8 0 10.6 10.6 10.6 10.6
TFAeti,' x lo3
M
10.7 10.7 10.6 10.6 8.6 0 8.6 0 8.0 8.0 8.6 0 8.6 0 8.6 0 8.6 0 13.1 13.1 13.1 13.1
TFAett/
X lo4,sec-L-,
-k
Mn(II1)
kMf
koa
km/ko
4.1 4.1 4.1 4.1 3.3
23.2 22.0 20.3 21.4 16.9 3.3 15.6 3.2 15.3 17.5 8.5 1.5 9.1 1.1 5.6 Ca. 1.0 5.2 Ca. 1.0 25.6 21 .o 21.0 25.0
19.0 19.0 19.0 19.0 15.2 4.3 15.2 5.0 15.2 14.0 15.2 4.3 15.2 5.0 15.2 4.3 15.2 5.0 19.0 19.0 19.0 19.0
1.2 1.2 1.1 1.1 1.1 0.8 1.0 0.7 1.1 1.2 0.56 0.34 0.60 0.21 0.37 0.22 0.34 0.20 1.4 1.1 1.1 1.3
3.3 3.0 3.0 3.3 3.3 3.3 3.3
5.0 5.0 5.0 5.0
a 2.60 X M Mn(II1). 99 wt pivalio acid-acetic acid. Molar ratio; added salt/Mn(III) = 1.0. Trifluoroacetic acid, TFA/ Mn(II1) = 5.0 when applicable. E Effective TFA concentration = initial TFA concentration - n[M(OAc),]. f Average of at least two Anhydrous. runs. 0 ko was determined from plots of kl us. TFA concentration (data in Table VII). h Dihydrate.
Table IX. Temperature Dependence of the Reduction of Mn(II1) in Pivalic Acid. Effect of Acetic Acid and Trifluoroacetic Acid HOAc,
M 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 3.5 3.5 3.5 3.5 3.5 3.5
CFr COOH, Temp, "C lo2 M
0 0 0 0 0
0 13.8 13.8 13.8 13.8 13.8 13.8 0 0
0 13.8 13.8 13.8
80 80 70 70 60 60 70 70
60 60 50 50 80 70 60 70
60 51
kl X lO4,a seC-1 8 . 9 f 0.6b 9.7 f0 . 5 ~ 4.3f0.2b 5.0 f 0 . 5 ~ 2 . 0 f 0.3b 2 . 1 z!z 0.10 28.9 f 0 . 2 ~ 2 7 & 10 1 6 . 7 f 1.4b 1 7 . 9 f 0.40 7 . 4 f 0.2b 7 . 7 f 0.4" 13.8 f 0.8b 3 . 0 f 0.2b 0 . 7 5 0.03b 22.8 f 1 . P 11.3 f 0 . 0 4.6 f 0 . 0
a Average of at least two runs. (OAC)~. Calculated at 70".
AH*, AS*,d kcal/mole eu
17.1
-24
14.3
-18
34.6
+26
16.7
-22
M ~ ( O A C ) ~ . ~ H5~MnO.
acetates as well as lead(I1) and cobalt(I1) acetates were also examined. None of these, with the exception of cobalt(I1) acetate, showed an effect comparable to manganese(I1) acetatess We tentatively ascribe the behavior of the latter to the formation of MnI". MnT1mixed valence complexes with rather low formaMnI"(OzCCH3)s
+ MnII(OzCCH& *
MnII1(O2CCH&Mn1I (12)
tion constants and extinction coefficients. The carboxylato ligands associated with these metal species, however, can act as bases and their effect in this capacity (9) (a) Changes in the absorption spectrum of MnlI1 in acetic acid solutions of alkali metal acetates at higher concentrations have been reported;' (b) W. J. de Klein, Thesis, Leiden, 1967.
cannot be e x c l ~ d e d . ~ Further !~ work is necessary to establish this point. Spectral Studies of the Disproportionation of Mn"' by Electron Spin Resonance. At low concentrations, the amplitudes of the first derivative of the hyperfine lines (at constant line width) of Mn" in aqueous solutions are proportional to the concentration.1° We found to 3 X M that standard solutions of 1 X Mn" acetate in acetic acid showed the same behavior quantitatively. M Mn"' in acetic The disproportionation of acid was assumed to be similar to that observed in aqueous acid solutions. The disproportionation constant Kd is then given by eq 13, where (Mnlll)o is the concentration of Mn"' charged. It was further
Kd =
(Mn11)2 [(Mnlll)o - 2(MnI1)l2
assumed that MnIV had no effect on the esr spectrum of Mn". Measurements of Mn" were made at various concentrations of Mn"' and temperatures. Reversibility in eq 1 was shown by the addition of Mn" and repeatedly recycling the temperature. Values of Kd obtained in this manner increased from 2.03 f 0.06 X at 30", 2.43 f 0.13 X l e 4 at 50" to 3.08 i= 0.24 X l e 4 at 70'. These values were essentially the same for anhydrous manganese(II1) acetate and its dihydrate in acetic acid solution^.^^^^ The value of Kd increased from 2.0 X l e 4 to 4.1 X l e 4 on the addition of trifluoroacetic acid (6.5 X l e 2M> at 44". (10) G. G. Guilbault and G. Lubrano, Anal. Lerfers, 1, 725 (1968); Anal. Chem., 41, 1100 (1969).
(11) Cf.also values in aqueous acid solutions: H. Diebler and N. Sutin, J . Phys. Chem., 68, 174 (1964); R. G. Selirn and J. J. Lingane, Anal. Chim. Acra, 21,536 (1959).
Anderson, Kochi / Manganese(III1 Complexes
2456 Discussion Manganese(II1) species oxidize carboxylic acids by two competing and largely independent paths : oxidative decarboxylation and alkyl oxidation. For example, the half-reactions for acetic acid are represented by eq 14 and 15, respectively. For a given carboxylic acid, oxidative decarboxylation CHaCOOH
isobutyric > pivalic acid. If this is correct the observed relative rates of decarboxylation must represent lower limits to rates of homolysis. These relative rates correlate with the stability of the alkyl radical [(CHB)BC* > (CH&CH. > CH3CHKH2.1 and suggest that the stretching of the alkyl-carbonyl bond in the transition state is simultaneous with the reduction of Mn"'. Multibond cleavage of the carboxylate
dioxide, and an alkyl radical. The rate of reduction of Mn"' follows first-order kinetics but is retarded by the MnII formed during the reaction. Kinetic and spectral studies allude to a mixed valence complex between Mn"' and Mn" carboxylates. The formation of these binuclear complexes is inhibited by strong acids which also accelerate the reduction of Mn'I' by formation of cationic Mn"' species. The latter also oxidize alkyl radicals more efficiently than their neutral counterparts. Neither Mn"' species can compete, however, with the efficient Cu(I1) scavenger which converts alkyl radicals clearly to alkenes. Pivalic acid undergoes catalytic decarboxylation in the presence of oxygen. The mechanism of this autoxidation follows directly from the mechanism of oxidative decarboxylation by Mn"'.
RCOzMnI1I-+[R..
Materials. Manganese(II1) acetate dihydrate was prepared from potassium permanganate and manganese(I1) acetate in glacial acetic acid.26 It was dried in vacuo over potassium hydroxide for 3 days at room temperature. The electron spin resonance spectra showed the presence of less than 2 mole % Mn". Anhydrous manganese(II1) acetate was prepared by a modified method6,14 and dried in the same manner after washing several times with anhydrous ether. The manganese(II1) content was determined by treatment with a known excess ferrous solution followed by back-titration with a standard ceric solution. The analyses are given in Table X and compared with other data. Manganese(II1) n-butyrate dihydrate was prepared by metathesis of manganese(II1) acetate dihydrate and n-butyric acid. The acetic acid was removed under vacuum. Anal. Calcd for C12HZ508Mn: 15.6% MnIII. Found: 15.1 Mn".
.CO~...Mnli---t R.
+ C 0 2+ MnI1
(40)
moiety is also observed in decarboxylations effected by c~~~~~~ and to a lesser extent PbIV oxidants.35 On the other hand, the facile thermal decarboxylations induced by Ag113?and photochemical decarboxylations by CeIV31show no selectivity. Acyloxy radicals have been proposed as intermediates which decarboxylate in a fast subsequent step. hv
+
RCOzCeIY +CeTT1 RC02. (1)
+ Re (2)
+ COZ
(41)
Multibond decarboxylations of acids by metal oxidants are associated with processes requiring a relatively high activation energy and derive driving force from the liberation of carbon dioxide. Stabilization of the alkyl radical also helps the concerted process, and t-alkyl radicals contribute the most and primary radicals the least toward lowering the barrier. 4 2 Oxidative decarboxylation must compete with the alkyl oxidation of acids (eq 15) which depends on the availability of a-hydrogens. Tertiary acids, thus, represent optimum examples for oxidative decarboxylation whereas acetic and other primary acids are constructed for alkyl oxidation. The results presented in Tables I, 11, and 111 are consistent with this trend. It is interesting to note that those oxidations (Ag" and photochemical CeIV49 which proceed via acyloxy radicals are much less sensitive to the competiting alkyl oxidations, and high efficiency is achieved in the decarboxylation process even with acetic acid. Summary. The oxidative decarboxylation of acids by Mn"' proceeds by multibond homolysis of a Mn"' carboxylate and simultaneously generates Mn", carbon (41) The foregoing statements can be reconciled if one considers that the effects of minor differences in rates (especially of preequilibria) are highly magnified in a competitive situation. The discrepancy is especially serious when the homolysis of the MnlI1 carboxylate is only slightly slower than the metathesis. The possibility that at least three carboxylato ligands in the MnlI1 monomer and six in the trimer" can exchange serially complicates the issue further, especially since each of these MnlI1 species probably decarboxylates at a different rate. (42) J. K. Kochi, J. D. Bacha, and T.W. Bethea, J . Am. Chem. SOC., 89, 6538 (1967). (43) The thermal decarboxylation of acids by CeIV is slowal except in the presence of strong acids, and recently Heiba and Dessa have shown that carboxymethyl radicals are formed in high yields from acetic acid. (We wish to thank these authors for kindly communicating these results to us prior to publication.)
Experimental Section
Table X. Elemental Analyses of MnIII Complexes -Calcd, MnIII Mn(OAc)s.2HzO Mn(0Ac)a Mn3(OAc)eOHa
%-C
H
-Found, Mn"'
%-C
H
20.5 26.87 4.85 20.9 27.34 4.95 23.67 31.05 3.91 23.6 30.12 4.10 25.3 29.36 3.82 25.2a 29.40"
= See ref 15.
All acids were purified by distillation. The Enjay Chemical Co. generously donated the pivalic acid. The esters, whether obtained from commercial sources or prepared by standard procedures, were distilled before use. The metal acetates and trifluoroacetates used in our studies were dried at 100" in uacuo for 24 hr. Pivalic acid-d and acetic acid-d were prepared as described previously.30 Product Studies. General Procedure. Approximately 2.7 mmoles of rnanganese(II1) acetate and other salts as required were weighed into a 50-1111 T round-bottom flask. A 10-ml sample of the appropriate carboxylic acid was added by pipet and the flask sealed with a gas-tight rubber septum. All rubber septa were washed with toluene and acetone to remove inhibitors. The flask was degassed in vacuo for 30 min. Acid catalysts when needed were dissolved in the carboxylic acid, degassed, and added to the reaction vessel uia a hypodermic syringe. The flask was then placed in a thermostated oil bath and the contents were stirred magnetically. The completion of the reaction in most cases was determined visually by a change in color from dark red-brown to colorless. Cupric acetate obscured the visual end point and these reactions were allowed to run for approximately one and one-half times as long as the uncatalyzed reaction to ensure completion. The oxygenbomb experiment was carried out using a procedure previously described." Analytical Procedure. When the reaction was complete, the reaction flask was immediately cooled in a carbon dioxide-isopropyl alcohol bath and known volumes of reference gases were added with a hypodermic syringe. The reaction flask was warmed
Anderson, Kochi / Manganese(III) Complexes
2460 Kinetic Procedure. Kinetics were followed on a Beckmann DB-G spectrophotometer equipped with a Sargent Model SRLG recorder. Temperature control was within i=0.4"in the thermostated compartment. The manganese(II1) acetate concentrations were determined spectrophotometrically at 462 nm (shoulder). Stock solutions of ca. 1.3 X M manganese(II1) acetate in pivalic acid were used and diluted by a factor of 5 with pivalic acid for the kinetic studies. Added acids and bases were dissolved in pivalic acid and added as required. After the reaction solution had been made up, a portion was transferred to a I-cm cell which was sealed with a rubber septum. The solution was then degassed with helium introduced through a hypodermic needle for 10-15 min. The cell was preheated with shaking to within 10' of the desired temperature before being placed in the spectrophotometer. Epr Method. All spectra were run on a Varian E-3 epr spectrometer. The same sample tube was used for the sample MnlI1 and standard MnlI solutions. The temperature was controlled to within 1.0'.
to room temperature, helium added to increase the pressure to slightly above atmospheric, and the flasks were vigorously shaken. The gases were then analyzed by gas chromatography. Mixtures of known volumes of products and reference gases were analyzed under reaction conditions to obtain calibration factors. After the gas analysis, a reference ester was added to the reaction solution and the solution brought to a volume of 50 cc with glacial acetic acid. A 15-ml aliquot of the solution was then added to 15 ml of diethyl ether. The ether solution was washed three times with water and three times with saturated sodium bicarbonate solution to remove the excess carboxylic acid. The ether solution was analyzed for ester products by gas chromatography. Solutions containing known concentrations of product and reference esters were analyzed in the same manner to obtain calibration factors. Gas chromatographic analyses of gaseous products were performed on instruments equipped with thermal conductivity detectors. Carbon dioxide was analyzed on a 2-ft Poropak Q column at room temperature using ethane as a reference gas. Propane, propene, isobutane, and isobutene were analyzed on a 15-ft 30% Dowtherm on firebrick column at room temperature using nbutane as a marker. Esters were analyzed on instruments employing flame-ionization detectors (Varian Aerograph, Model 1200, and Varian Aerograph, Model 200). An 8-ft DEGS column was used to analyze n-propyl n-butyrate, isopropyl n-butyrate, isopropyl isobutyrate, and isopropyl acetate with n-butyl acetate as the marker. A 6-ft Morflex column was used to analyze t-butyl pivalate and t-butyl acetate with n-butyl acetate as the marker.
Acknowledgment. We wish to thank the National Science Foundation for generous financial support of this work, Dr. Sheldon Lande for the study of the autoxidative decarboxylations catalyzed by MnI'I, and Professor E. C . Kooyman for kindly sending us copies of the theses of Drs. W. J. DeKlein, R. E. Van der Ploeg, R. W. de Korte, and L. W. Hessel.
Organoboranes. IX. Structure of the Organoboranes Formed in the Reaction of 1,3-Butadiene and Diborane in the Stoichiometric Ratio. An Unusual Thermal Isomerization of These Organoboranes Herbert C. Brown, Ei-ichi Negishi,' and Shyam K. Gupta'
Contributionfrom the Richard B. Wetherill Laboratory of Purdue University, Lafayette, Indiana 47907. Received August 5, 1969 Abstract: The hydroboration of 1,3-butadiene with diborane in the stoichiometric ratio a t 0' produces predominantly the dumbbell-shaped structures, 1,3- a n d 1,Cbis(l-boracyclopenty1)butanes. Although the ratio of the two major isomers varies somewhat with the reaction conditions, the 1,3 isomer is the preferred product, -70%, under kinetically controlled conditions. Under isomerization conditions (140-170"), this isomer rapidly vanishes and the product is converted into 75% 1,l- and 25% 1,4-bis(l-boracyclopentyl)butanes. Independent syntheses of 1,land 1,4-bis(l-boracyclopentyl)butanes was achieved by treating 1-butyne and 1,3-butadiene with B-methoxyboracyclopentane and lithium aluminum hydride. The structure of the 1,4 isomer was confirmed by treatment with carbon monoxide to form 1,l '4etramethylenedicyclopentanol. In marked contrast t o the behavior of simple trialkylboranes, the boracyclopentane ring is readily opened by solvolysis with water or alcohol. Deuterolysis with heavy water, followed by oxidation, provides a convenient synthesis of 1-butanol-CdL. T h e present understanding of the reaction products a n d the chemistry of "bisborolane" permits postulation of a reaction mechanism for the hydroboration of 1,3-butadiene.
he cyclic h y d r o b o r a t i o n of dienes with thexylT b o r ane followed by carbonylation provides a convenient new synthetic route to cyclic and polycyclic ketones. 3,4 Hydroboration with diborane, followed by carbonylation, has been successfully applied
to trieness and to one d i e n e , 1,5-~yclooctadiene.~However, extension of this p r o m i s i n g new synthetic a p p r o a c h to dienes generally has been handicapped by the confused state of our knowledge as to the structure of the products formed from the reaction of diborane with dienes.?
(1) Postdoctorate Research Associate on a research grant, DA 31-124 ARO(D) 453, supported by the U. S. Army Research Office (Durham). (2) H. C. Brown and C. D. Pfaffenberger, J . Amer. Chem. Soc., 89,
5475 (1967). (3) H. C. Brown and E. Negishi, ibid., 89, 5477 (1967); H. C. Brown and E. Negishi, Chem. Commun., 594 (1968). (4) For a general review of this development with complete literature references, see H. C. Brown, Accounts Chem. Res., 2,65 (1969).
(5) H. C. Brown and E. Negishi, J . Amer. Chem. SOC.,89, 5478 (1967); H. C. Brown and W. C. Dickason, ibid., 91, 1226 (1969); H. C. Brown and E. Negishi, ibid., 91, 1224 (1969). (6) E. F. Knights and H. C. Brown, ibid., 90,5280, 5281, 5283 (1968). (7) The hydroboration-oxidation of dienes has been subjected to de-
tailed study. However, for such production of enols or diols, the precise structure of the boron intermediates is not of major importance and
Journal of the American Chemical Society 1 92:8 1 April 22, 1970
